ACKO Expert: AI-assisted support

Working notes

From the project brief. Draft wording, to be rewritten with Mobin.

Designed an LLM-powered support experience around existing customer context and lifecycle needs, with an explicit route to human assistance when the bot could not resolve a request.
Conversational UX with a content designer, adaptive response patterns, deep links to self-service, a Contact Us redesign, chatbot UI components and improvements to an internal monitoring experience.
A related health insurance journey concept influenced a later redesign.
Mobin supplied working metrics (call volume, estimated savings, bot upvotes, a related sales uplift). They stay out until ownership and definitions are confirmed.

To write with Mobin: what was broken or missing, for whom, and why it mattered then.

To write with Mobin: what users needed, what the business needed, and where those pulled against each other.

To write with Mobin: what Mobin owned, who he worked with, and what was out of scope.

To write with Mobin: the constraints that shaped the work and the evidence behind key decisions.

To write with Mobin: directions considered, the one chosen, and why the others were rejected.

To write with Mobin: the end-to-end flow and the interaction details that carry it.

To write with Mobin: how the design was tested, what changed, and what was traded away.

To write with Mobin: verified results, what each metric means and how it was measured.

To write with Mobin: what Mobin learned, what he would change, and what remains unknown.